Setup python (#10)

* Setup python

.gitignore

@@ -1,9 +1,24 @@
 bazel-*
+build
+mediapipe.egg-info
+mediapipe/__pycache__/
 mediapipe/MediaPipe.xcodeproj
 mediapipe/MediaPipe.tulsiproj/*.tulsiconf-user
+mediapipe/models/ovms/face_detection_short_range/
+mediapipe/models/ovms/face_landmark/
+mediapipe/models/ovms/hand_landmark_full/
+mediapipe/models/ovms/hand_recrop/
+mediapipe/models/ovms/iris_landmark/
+mediapipe/models/ovms/palm_detection_full/
+mediapipe/models/ovms/pose_detection/
+mediapipe/models/ovms/pose_landmark_full/
+mediapipe/models/ovms/ssdlite_object_detection/
+mediapipe/models/ssdlite_object_detection_labelmap.txt
 mediapipe/provisioning_profile.mobileprovision
+mediapipe/python/__pycache__/
 node_modules/
 .configure.bazelrc
 .user.bazelrc
 .vscode/
 .vs/
+*.mp4

Dockerfile.openvino

@@ -85,12 +85,6 @@ RUN apt-get update && apt-get install --no-install-recommends -y \
     libopencv-imgproc-dev \
     libopencv-video-dev \
     build-essential \
-    libboost-filesystem1.71.0 \
-    libboost-thread1.71.0 \
-    libboost-program-options1.71.0 \
-    libboost-chrono1.71.0 \
-    libboost-date-time1.71.0 \
-    libboost-atomic1.71.0 \
     libjson-c4 \
     unzip
 

WORKSPACE

@@ -638,10 +638,12 @@ http_archive(
     build_file = "@//third_party:halide.BUILD",
 )
 
+load("@bazel_tools//tools/build_defs/repo:git.bzl", "git_repository")
+
 git_repository(
     name = "ovms",
     remote = "https://github.com/openvinotoolkit/model_server",
-    commit = "77c30dc3f153b3ee78336a3a75c09af4e23c14a4", # MP update to 10.3 in OVMS
+    commit = "ad1381fde838f2ac2d23117df78c186a96134fcc", # Fix azure patch (#2107)
 )
 
 # DEV ovms - adjust local repository path for build

mediapipe/__init__.py

@@ -1,13 +1,13 @@
-# Copyright 2019 - 2022 The MediaPipe Authors.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#      http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
+
+from mediapipe.python import *
+import mediapipe.python.solutions as solutions 
+import mediapipe.tasks.python as tasks
+
+
+del framework
+del gpu
+del modules
+del python
+del mediapipe
+del util
+__version__ = '1.0'

mediapipe/examples/python/README.md

@@ -0,0 +1,33 @@
+# OVMS python examples
+- Building docker container with dependencies
+```bash
+git clone https://github.com/openvinotoolkit/mediapipe.git
+cd mediapipe
+make docker_build
+```
+
+- Start the container
+```bash
+docker run -it mediapipe_ovms:latest bash
+```
+
+- Prepare models for ovms
+```bash
+python setup_ovms.py --get_models'
+```
+
+- Build and install mediapipe python package
+Make sure you are in /mediapipe dirctory
+Below command takes around 1 hour depending on your internet speed and cpu
+```bash
+pip install .
+```
+
+- Run example ovms python script
+```bash
+python build/lib.linux-x86_64-cpython-38/mediapipe/examples/python/ovms_object_detection.py
+```
+
+- This script will run object detection on input video, as described in this c++ example
+[OVMS Object Detection](../desktop/object_detection/README.md)
+[Original demo documentation](https://google.github.io/mediapipe/solutions/object_detection)

mediapipe/examples/python/ovms_object_detection.py

@@ -0,0 +1,21 @@
+# Copyright (c) 2023 Intel Corporation
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+import mediapipe as mp
+ovms_object_detection = mp.solutions.ovms_object_detection
+with ovms_object_detection.OvmsObjectDetection(side_inputs=
+        {'input_video_path':'/mediapipe/mediapipe/examples/desktop/object_detection/test_video.mp4',
+         'output_video_path':'/mediapipe/tested_video.mp4'}) as ovms_object_detection:
+        results = ovms_object_detection.process()

mediapipe/modules/ovms_modules/BUILD

@@ -0,0 +1,39 @@
+# Copyright (c) 2023 Intel Corporation
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+load(
+    "//mediapipe/framework/tool:mediapipe_graph.bzl",
+    "mediapipe_simple_subgraph",
+)
+load(
+    "//mediapipe/framework/tool:mediapipe_files.bzl",
+    "mediapipe_files",
+)
+load("//mediapipe/framework/port:build_config.bzl", "mediapipe_proto_library")
+load("//mediapipe/framework:mediapipe_cc_test.bzl", "mediapipe_cc_test")
+
+licenses(["notice"])
+
+package(default_visibility = ["//visibility:public"])
+
+mediapipe_simple_subgraph(
+    name = "object_detection_ovms",
+    graph = "object_detection_ovms.pbtxt",
+    register_as = "ObjectDetectionOvms",
+    deps = [
+        "//mediapipe/graphs/object_detection:desktop_ovms_calculators",
+        "@ovms//src:ovms_lib",
+    ],
+)

mediapipe/modules/ovms_modules/object_detection_ovms.pbtxt

@@ -0,0 +1,206 @@
+# MediaPipe graph that performs object detection on desktop with OpenVINO Model Server
+# on CPU.
+# Used in the example in
+# mediapipe/examples/desktop/object_detection:object_detection_openvino.
+
+# max_queue_size limits the number of packets enqueued on any input stream
+# by throttling inputs to the graph. This makes the graph only process one
+# frame per time.
+max_queue_size: 1
+
+# Decodes an input video file into images and a video header.
+node {
+  calculator: "OpenCvVideoDecoderCalculator"
+  input_side_packet: "INPUT_FILE_PATH:input_video_path"
+  output_stream: "VIDEO:input_video"
+  output_stream: "VIDEO_PRESTREAM:input_video_header"
+}
+
+# Transforms the input image on CPU to a 320x320 image. To scale the image, by
+# default it uses the STRETCH scale mode that maps the entire input image to the
+# entire transformed image. As a result, image aspect ratio may be changed and
+# objects in the image may be deformed (stretched or squeezed), but the object
+# detection model used in this graph is agnostic to that deformation.
+node: {
+  calculator: "ImageTransformationCalculator"
+  input_stream: "IMAGE:input_video"
+  output_stream: "IMAGE:transformed_input_video"
+  node_options: {
+    [type.googleapis.com/mediapipe.ImageTransformationCalculatorOptions] {
+      output_width: 320
+      output_height: 320
+    }
+  }
+}
+
+# Converts the transformed input image on CPU into an image tensor as a
+# OpenVINOTensor. The zero_center option is set to true to normalize the
+# pixel values to [-1.f, 1.f] as opposed to [0.f, 1.f].
+node {
+  calculator: "OpenVINOConverterCalculator"
+  input_stream: "IMAGE:transformed_input_video"
+  output_stream: "TENSORS:image_tensor"
+  node_options: {
+    [type.googleapis.com/mediapipe.OpenVINOConverterCalculatorOptions] {
+      enable_normalization: true
+      zero_center: true
+    }
+  }
+}
+
+# Runs a TensorFlow Lite model on CPU that takes an image tensor and outputs a
+# vector of tensors representing, for instance, detection boxes/keypoints and
+# scores.
+node {
+  calculator: "OpenVINOModelServerSessionCalculator"
+  output_side_packet: "SESSION:session"
+  node_options: {
+    [type.googleapis.com / mediapipe.OpenVINOModelServerSessionCalculatorOptions]: {
+      servable_name: "ssdlite_object_detection"  # servable name inside OVMS
+      servable_version: "1"
+      server_config: "mediapipe/calculators/ovms/config.json"
+    }
+  }
+}
+node {
+  calculator: "OpenVINOInferenceCalculator"
+  input_side_packet: "SESSION:session"
+  input_stream: "OVTENSORS:image_tensor"
+  output_stream: "OVTENSORS2:detection_tensors"
+  node_options: {
+    [type.googleapis.com / mediapipe.OpenVINOInferenceCalculatorOptions]: {
+          input_order_list :["normalized_input_image_tensor"]
+          output_order_list :["raw_outputs/box_encodings","raw_outputs/class_predictions"]
+          tag_to_input_tensor_names {
+            key: "OVTENSORS"
+            value: "normalized_input_image_tensor"
+          }
+          tag_to_output_tensor_names {
+            key: "OVTENSORS1"
+            value: "raw_outputs/box_encodings"
+          }
+          tag_to_output_tensor_names {
+            key: "OVTENSORS2"
+            value: "raw_outputs/class_predictions"
+          }
+        }
+  }
+}
+
+# Generates a single side packet containing a vector of SSD anchors based on
+# the specification in the options.
+node {
+  calculator: "SsdAnchorsCalculator"
+  output_side_packet: "anchors"
+  node_options: {
+    [type.googleapis.com/mediapipe.SsdAnchorsCalculatorOptions] {
+      num_layers: 6
+      min_scale: 0.2
+      max_scale: 0.95
+      input_size_height: 320
+      input_size_width: 320
+      anchor_offset_x: 0.5
+      anchor_offset_y: 0.5
+      strides: 16
+      strides: 32
+      strides: 64
+      strides: 128
+      strides: 256
+      strides: 512
+      aspect_ratios: 1.0
+      aspect_ratios: 2.0
+      aspect_ratios: 0.5
+      aspect_ratios: 3.0
+      aspect_ratios: 0.3333
+      reduce_boxes_in_lowest_layer: true
+    }
+  }
+}
+
+# Decodes the detection tensors generated by the TensorFlow Lite model, based on
+# the SSD anchors and the specification in the options, into a vector of
+# detections. Each detection describes a detected object.
+node {
+  calculator: "OpenVINOTensorsToDetectionsCalculator"
+  input_stream: "TENSORS:detection_tensors"
+  input_side_packet: "ANCHORS:anchors"
+  output_stream: "DETECTIONS:detections"
+  node_options: {
+    [type.googleapis.com/mediapipe.OpenVINOTensorsToDetectionsCalculatorOptions] {
+      num_classes: 91
+      num_boxes: 2034
+      num_coords: 4
+      ignore_classes: 0
+      apply_exponential_on_box_size: true
+
+      x_scale: 10.0
+      y_scale: 10.0
+      h_scale: 5.0
+      w_scale: 5.0
+    }
+  }
+}
+
+# Performs non-max suppression to remove excessive detections.
+node {
+  calculator: "NonMaxSuppressionCalculator"
+  input_stream: "detections"
+  output_stream: "filtered_detections"
+  node_options: {
+    [type.googleapis.com/mediapipe.NonMaxSuppressionCalculatorOptions] {
+      min_suppression_threshold: 0.4
+      min_score_threshold: 0.6
+      max_num_detections: 5
+      overlap_type: INTERSECTION_OVER_UNION
+    }
+  }
+}
+
+# Maps detection label IDs to the corresponding label text. The label map is
+# provided in the label_map_path option.
+node {
+  calculator: "DetectionLabelIdToTextCalculator"
+  input_stream: "filtered_detections"
+  output_stream: "output_detections"
+  node_options: {
+    [type.googleapis.com/mediapipe.DetectionLabelIdToTextCalculatorOptions] {
+      label_map_path: "/mediapipe/mediapipe/models/ssdlite_object_detection_labelmap.txt"
+    }
+  }
+}
+
+# Converts the detections to drawing primitives for annotation overlay.
+node {
+  calculator: "DetectionsToRenderDataCalculator"
+  input_stream: "DETECTIONS:output_detections"
+  output_stream: "RENDER_DATA:render_data"
+  node_options: {
+    [type.googleapis.com/mediapipe.DetectionsToRenderDataCalculatorOptions] {
+      thickness: 4.0
+      color { r: 255 g: 0 b: 0 }
+    }
+  }
+}
+
+# Draws annotations and overlays them on top of the input images.
+node {
+  calculator: "AnnotationOverlayCalculator"
+  input_stream: "IMAGE:input_video"
+  input_stream: "render_data"
+  output_stream: "IMAGE:output_video"
+}
+
+# Encodes the annotated images into a video file, adopting properties specified
+# in the input video header, e.g., video framerate.
+node {
+  calculator: "OpenCvVideoEncoderCalculator"
+  input_stream: "VIDEO:output_video"
+  input_stream: "VIDEO_PRESTREAM:input_video_header"
+  input_side_packet: "OUTPUT_FILE_PATH:output_video_path"
+  node_options: {
+    [type.googleapis.com/mediapipe.OpenCvVideoEncoderCalculatorOptions]: {
+      codec: "avc1"
+      video_format: "mp4"
+    }
+  }
+}

mediapipe/python/BUILD

@@ -58,6 +58,8 @@ pybind_extension(
         "//mediapipe/framework/formats:rect_registration",
         "//mediapipe/modules/objectron/calculators:annotation_registration",
         "//mediapipe/tasks/cc/vision/face_geometry/proto:face_geometry_registration",
+        # OVMS lib
+        "@ovms//src:ovms_lib",
     ],
 )
 

mediapipe/python/solutions/__init__.py

@@ -23,5 +23,6 @@
 import mediapipe.python.solutions.hands_connections
 import mediapipe.python.solutions.holistic
 import mediapipe.python.solutions.objectron
+import mediapipe.python.solutions.ovms_object_detection
 import mediapipe.python.solutions.pose
 import mediapipe.python.solutions.selfie_segmentation